The issue is about a buck converter design using LTC3895. The reference design for LTC3895 is followed to regulate from 48V to 36V. The schematic for the output circuitry is shown Fig. 1. The buck converter output is to a min 3.6 ohm load for 10A max current. The BG voltage is monitored and shown at Fig. 2. During start up, the BG voltage stays at 10V and there is no over-voltage detected. But over 3 cycles, the BG voltage degrades and pulled down to 1V (2V/div vertically), indicating a damage to the BG output. The power to the board is turned off and the resistance is measured between BG and GND. Resistance measurement shows 0 ohm between the BG and GND pins, indicating a short.
Questions:
1. What are the design concerns for LTC3895 design I need to be aware of for this case?
2. What are the recommendations for the design?
3. Any similar cases reported?
4. Any comments on the BG voltage degradation over initial transient?
Fig. 1 LTC3895 output circuitry
Fig. 2 LTC3895 grounding scheme
Fig. 3 BG voltage
